Structure of aqueous glutaraldehyde - ACS Publications

51240-39-2; 8, 13861-75-1; 9a, 51240-40-5; 9b, 51240-41-6; lOaA,. 30671-33-1; lObA, .... Union Carbide Corporation, Tarrytown, New York 10591. Receive...
4 downloads 0 Views 371KB Size
1666 J . Org. Chem., Vol. 39, No. 12, 1974

Whipple and Ruta

6.17 g ( m ) ; uv Amax ( E t O H ) 253 nm (e %lo), 284 (1210); pmr 6 6.88(dd, 1, J = 2, 7 H z , H-121, 7.02 (dt, 1, J = 2, 7 Hz. HG-C

lo),7.15(dt, 1, J = 2, 7 Hz,H-11). Anal C a l c d f o r C I ~ H ~ ~ O ~ T C, L T 71.82; ~: H.7.09; N, 9.85.F o u n d : C,71.61:H,7.19;N,10.03. A s o l u t i o n of 110 m g of lOdA in 10 ml o f 10% aqueous acetic a c i d was r e f l u x e d f o r 20 hr a n d t h e n cooled. n e u t r a l i z e d w i t h s o l i d p o t a s s i u m bicarbonate, a n d e x t r a c t e d w i t h m e t h y l e n e chloride. T h e e x t r a c t was d r i e d a n d evaporated. T h e pmr s p e c t r u m of t h e o i l y residue, 102 m g , showed the presence of a 1.6:1m i x t u r e of lOdA a n d IOdB, respectively. A s o l u t i o n of 100 m g of lOdB in 10 ml of 10% aqueous acetic a c i d was refluxed for 20 hr. S i m i l a r r e a c t i o n w o r k - u p l e d to 77 m g of p r o d u c t whose t l c (silica gel, 1.6:1chloroform-methanol) revealed iOdA a n d IOdB, t h e former p r e d o m i n a t i n g , a n d whose pmr s p e c t r u m e x h i b i t e d signals o f a 1.9:1 m i x t u r e of lOdA a n d lOdB, respectively.

Registry vo.-3b, 4695-82-3;4c, 2671-38-7;5, 51240-38-1;6b, 51240-39-2;8, 13861-75-1;9a, 51240-40-5;9b, 51240-41-6:iOaA, 30671-33-1; lObA, 51268-47-4; lOcA, 51240-42-7; lOdA, 51240-43-8; lObB, 51268-48-5;lOdB, 51268-49-6;t r y p t o p h o l , 526-55-6;k e t o chloroindolenine, 51240-44-9.

References and Notes This investigation was supported by the U. S Public Health Service. Part XII: E, Wenkert, P. W. Sprague, and R. L. Webb, J. Org. Chem., 38, 4305 (1973). U . S. Public Health Service Predoctoral Fellow, 1967-1971. E Wenkert, Accounts Chem. Res., 1, 78 (1968) T. Ohnuma, T. Oishi, and Y. Ban. J. Chem. SOC., Chem. Commun., 301 (1973). and references cited therein. W. 8 . Lawson and B. Witkop, J . Org. Chem., 26, 263 (1961). F. J McEvoy and G . R. Alien Jr , J . Org. Chem., 38,3350 (1973) P. L. Juiian, A . Magnani, J Piki, and W. Karpel, J. Amer. Chem. S o c , 70, 174 (1948); E. Wenkert and E. C Blossey, J. Org. Chem., 27, 4656 (1962) D G. Markees and A Burger, J. Amer. Chem. SOC., 71, 2031 (1949); E. Wenkert and T. L. Reid, Chem. ind. ( L o n d o n ) , 1390 (1953). i t ais0 was the product of attempted mesylation of 5 in pyridine. E. Wenkert, K G . Dave, and F . Hagiid, J . Amer. Chem. Soc., 87, 5461 (1965) M. Ohno, T. F. Spande. and 6. Witkop, J . Amer. Chem. SOC., 92, 343 (1970) N . Finch and W. I . Taylor. J . Amer. Chem. SOC..84, 3871 (1962). E Wenkert and J S. Bindra. unpublished observations.

(15) A . J Gaskell, H.-E. Radunz, and E , Winterfeidt, Tetrahedron, 26, 5353 (1970). (16) E. Wenkert and 8 . Wickberg, J. Amer. Chem. SOC., 87, 1580 (1965) (17) T. Nozoe, Chem. Pharm. Bull., 6 , 300 (1958); J. C. Seaton, M. D. Nair, 0. E. Edwards, and L. Marion, Can. J. Chem., 38, 1035 (1960); N. Finch and W. i . Taylor, J. Amer. Chem. Soc., 84, 1318 (1962); G . M. Badger, L. M. Jackman, R . F. Skiar, and E Wenkert, Proc. Chem. SOC.,206 (1963) (18) E . Wenkert, J . H. Udelhofen, and N . K. Bhattacharyya, J. Amer. Chem. SOC., 81, 3763 (1959) (19) W. F . Trager, C. M. Lee, J. D . Phiilipson, R . E. Haddock, D. Dwuma-Badu, and A . H. Beckett, Tetrahedron, 24, 523 (1968), and references cited therein to earlier work. (20) The indolizidine portion of l O a A and lOaB Is portrayed usually with C ( 5 ) , C ( 6 ) , C(7), and C(3) in coplanar r e l a t i ~ n s h l pSince . ~ ~ ~this ~~ conformation exposes the geminal C(7) substituents to an energetically unfavorable, eclipsing interactlon with the C(6) hydrogens, coplanarity between Nb, C(5), C ( 6 ) , and C(7) is preferred. This latter representation has been illustrated previously without e ~ p i a n a f i o n . ’ ~ (21) All alkaloids have been assumed to possess a trans-indoiizidlne structure in all solutions on the basis of a CISstructure being destabilized by extra 1,3-diaxiai interactions in the piperidine moiety.ig Whiie infrared and pmr spectral analyses of simply substituted indolizidines indicate the basic skeleton to be trans fused [H.S. Aaron and C P. Ferguson, Tetrahedron Lett., 6191 (1968); T. A Crabb, R . F Newton. and D. Jackson, Chem. Rev., 71, 109 (1971)], the difference of energy between the CIS and trans forms is too low and the steric as well as polar interactions of the cxindole ring with the indolizidine nucleus too subtle and complex to permit acceptance of a trans configuration for the latter in compounds 7 without rigorous proof. Equilibration in pyridine leading to l O a A has been assumed to be the consequence of electron repulsion between the carbonyl oxygen and Nb, while preference of lOaB in acetic acid has been ascribed to hydrogen bridging of a Nb-protonated species with the carbonyl ~ x y g e n . However ~ ~ , ~ ~this explanation appears to be oversimplified in view of the fact that the isomer ratio rarely ever exceeds 3:1, Indicating a small energy difference, and that substances 7 can be expected to be only minimally protonated in aqueous acetic acid Merely differences in solvation suffice to give variations of equilibrium positions. (22) The same configurations were assigned earlierT5on the basis of pmr data on 7a with assumed trans-indolizidine structures. (23) For convenience the oxindole alkaloid nomenclature is being maintained for the modeis 7. (24) E. Wenkert, J. S Bindra, C.-J. Chang, D. W. Cochran, and F. M . Scheli. Accounts Chem. Res.. 7, 46 (1974) (25) E. Wenkert, D. W. Cochran, E. W. Hagaman, F. M . Scheil, N. Neuss, A. S. Katner, P. Potier, C. Kan, M. Plat, M , Koch, H . Mehri, J. Poisson, N . Kunesch, and Y . Rolland, J. Amer. Chem. SOC., 95, 4990 (1973). (26) E. Wenkert, K. G . Dave, C. T. Gnewuch, and P. W Sprague, J. Amer. Chem. SOC.,90, 5251 (1968).

Structure of Aqueous Glutaraldehyde Earl B. Whipple” and Michael Ruta Cnion Carbide Corporation, Tarr)tobn, N e u ~York 10591

Recewed September 28,1973 T h e carbon-13 nmr s p e c t r u m of aqueous 25% g l u t a r a l d e h y d e has been assigned t o i n d i v i d u a l components in a n e q u i l i b r i u m m i x t u r e . T h e s o l u t i o n i s shown t o consist p r i m a r i l y of t h e cyclic hemiacetal, e q u a l l y present in i t s t w o isomeric forms, in e q u i l i b r i u m w i t h t h e free aldehyde. T h e r a t i o of these components varies strongly w i t h t e m p e r a t u r e . A p p r o x i m a t e l y 25% o f t h e m i x t u r e i s present as t h e l i n e a r h e m i h y d r a t e a n d t h e d i h y d r a t e in a b o u t a 2:l r a t i o , t h i s f r a c t i o n b e i n g m u c h less t e m p e r a t u r e dependent. H i g h e r order oligomers c o n t r i b u t e v e r y little t o the equilibrium mixture.

The structure of gluataraldehyde (pentanedial) in aqueous solution, in which form it is available as a commercial product,l has been the subject of several proton magnetic resonance ~ t u d i e s , ~ not - ~ all of which give consistent results. The currently accepted s t r ~ c t u r e .based ~ on the work by Hardy, Xicholls, and Rydon.3 is that the solution contains roughly equal amounts of the hemihydrate 11, the dihydrate 111, and the cyclic hemiacetai IV. It has not been possible t o distinguish individual structures in the proton magnetic resonance spectrum owing to the overlapping of complex bands. Carbon-13 nmr is much less subject to these restrictions. With broad band proton decoupling, I. 111, and N should give simple three-line,

CHO

I

CH?

HC(OH), ---+ H,O

1

CH2

I

CH,

I

+-

CH2

CH?

I

CH*

CHO

CHO

I

I1

I

I

I

HC(OHh H,O

---+

1

CHP

I

CH,

I CH2 I

HC(OH), I11

-

HOCH

C-

“O

H !(q

I I C I

(1) (2)

CH, 0 H d

HOCH

(3) (-1)

(5)

Tv

2:2:1 intensity patterns, and five equally intense lines should result from 11. Higher order oligomers or condensa-

J. Org. Chem., VoE. 39, No. 12, 1974 1667

Structure of Aqueous Glutaraldehyde

Table I Carbon-13 Line Positions and Assignments in Aqeous Glutaraldehyde

,-------c

aband--p---

Shift'

Linet1

14.6 16.9 17.1 17.3 19.4 20.25

1 2

3 4 5 6

a

r

_

~

_

__-_-_-

CzL and Cb band-------

C1 and Ca band-------

AssignmentC

Line

Shift

Assignment

Line

Shift

I IVb I1 I11 IVa

7 8 9 10 11 12 13 14

29.4 30.9 31.65

?

IVb IVa

15 16

91.35 91.6 92.3 95.3 208.4 209.1

17

? 11 (2)

36.7 37.1 42.5 43.0

18 19 20

Assignment

I11 I1

IVb IVa I 11 ( 5 )

I11 I 11 (4) Figure 1. b Parts per million downfield from external TMS. Internal dioxane occurs a t 66.4 ppm on this scale. See text. ?

- J' L.1.

C, 5

c2,4

c3

Figure 1. Carbon-13 nmr spectrum (exclusing aldehyde lines) of aqueous 25% glutaraldehyde a t 23" (22.63 MHz, increasing frequency from right to left). Line assignments are given in Table I. tion products should give more complex patterns than any of these, and therefore be fairly easy to recognize. There is also no possibility of solvent interference. The 22.63-MHz, proton-decoupled. naturally abundant carbon-13 spectrum of an aqueous, 25% solution of repurified (saturate commercial solution with NaC1, extract with ether, dry, strip, and distil3) glutaraldehyde is shown in Figure 1, and the line assignments are detailed in Table L5 The commercial material shows substantially the same spectrum. The assignments are based on chemical shift additivity rules6 and changes in the line intensities that occur as the result of variation in concentration and temperature. The two major components in solution each give rise to a 2:2:1 pattern whose chemical shifts all fall far outside the range of free aldehyde groups. While this would be generally consistent with Hardy, Nicholls, and Rydon's conclusion that I11 and IV are present in large quantity, we believe that these patterns instead represent the two geometrical isomers of IF', i. e., IVa and IVb.

HO

OH

OH

dH

IVb (trans) This is based on three main lines of argument. The first is that conformational energy considerations (anomeric effects) would lead one to expect both isomers to be present in comparable amounts of aqueous solution, while the driving force for the formation of 111 is much less apparent. The second is that the relative chemical shifts between the two observed species, particularly the - 2 . 5 ppm difference a t carbons 1 and 3 us. the smaller differIVa(cis)

- - - ~..

I

v

2000

Hz (220 MHz)

1000

Ho

-

~

-

0

Figure 2. 220-MHz proton magnetic resonance spectrum of purified glutaraldehyde (25% in DzO). ence of 0.7 ppm a t Cz, are in accord with the well-known y effect in cyclic systems.? Thirdly, the relative intensities of the two principal patterns do not change significantly between 26 and 50% aqueous glutaraldehyde, although the molarity of water decreases by approximately 10%. The equilibrium between I11 and N ,which involves an additional molecule of water, should be affected, while that between IVa and IVb would not. The assignment of the principal low-temperature structures to the two isomers of the cyclic hemiacetal, IVa, and IVb, is confirmed by the 220-MHz proton spectrum, Figure 2. Two strong, equally intense bands occur in the central region characteristic of -CH(OR)z groups, both of which show a principal doublet structure ( J = 6, J' = 2 Hz and J = 9.2, J = 2 Hz). The dihydrate would have to show triplet fine structure since the two adjacent methylene protons are isochronous and strongly coupled. IVa should exist in the diequatorial form, so that the anomeric proton would experience one strong (Jaa N 10 Hz) and one weak (Jae 2-4 Hz) vicinal coupling. Since axial protons generally occur a t higher field than equatorial, this would correspond to the upper band, which has the larger doublet splitting. Isomer IVb, which gives the lower field band, is the weighted average of two interconverting conformations, and should have vicinal coupling constants equal to l/Z(Jaa Jee) and '/Z(Jae $. J e a ) , respectively. The larger splitting is accordingly reduced by 30-40%, as is observed, and the smaller coupling is not substantially affected.

+

1668 J. Org. Chem., Vol. 39, No. 12, 1974

Whipple and Ruta

Table I1 Predicted0 Chemical Shifts of Glutaraldehyde Hydrates

----

Carbon number--

7

Compd

1

2

3

4

5

n-Pentaneb I I1 I11 Pyrand IVa IVb

13.7 202.OC 110.3 110.3 69.7

22.6 51.6 41.0 43.0 27.9 34.2 33.5

34.5 34.5 22.9 11.3 25.1 22.9 17.2

22.6 51.6 53.6 43.0 27.9 34.2 33.5

13.7 202.0~ 202.00 110.3 69.7 111.8 105.4

111.8

105.4

Based on reference compounds listed and additive corrections (ref 6) a

a

P

r

a

48.3 10.2 -5.8 31 -2

-OH -CHO

P

r

8

-OH(e) 43.2 7 . 9 - 1 . 1 - 1 . 6 -OH(a) 37.8 5 . 5 - 6 . 8 -0.7

for linear and cyclic derivatives, respectively. D. M. Grant and E. G. Paul, J . Amer. Chern. SOC.,86, 2984 (1964). n-Hexyl aldehyde value, converted t o TMS reference: J. B. Stothers and P. C. Lauterbur, Can. J . Chem., 42, 1563 (1964). d G. E. Maciel and G. B. Savitsky, J . Phys. Chem., 69, 3925 (1965). 70

-

60

!i r?

40-

IVa

a,

3

30-

C

that the chemical shifts between the adjacent methylene protons be large with respect to the geminal coupling constant between them in order to eliminate virtual coupling effects. A secondary set of five equally intense carbon-13 lines occurs that can be assigned to the hemihydrate 11. It is considerably less abundant than previously auserted, and each of its lines is accompanied by a weaker, secondary component. We assign these weaker lines to a mixture of the dihydrate I11 and the free aldehyde I. The chemical shifts, while not accurately predicted by additive substituent rules (Table II), do show consistent internal trends. The lines due to the free aldehyde could be positively identified, since its concentration has been shown to increase with elevation of the t e m p e r a t ~ r e .Three ~ weak, somewhat broader bands remain unassigned, and are probably due to oligomers. Based on the peak heights of lines 8, 9, 11, 12, and 13 of the carbon-13 spectrum, we estimate the following composition for purified aqueous glutaraldehyde in neutral solution a t 23": I, 4%; 11, 16%, 111, 9%; IVa, 35%; N b , 36%. This varies slightly with concentration and pH, and strongly with temperature. The approximate composition of commercial 25% aqueous glutaraldehyde1 is plotted as a function of temperature in Figure 3. Also shown is the fraction of free aldehyde groups calculated from previously available pmr data,4 and the corresponding points (x) that are determined from the composition as estimated by carbon-13; Le., y2 I1 + I. The agreement between the independent measurements and samples, particularly at low temperatures, is reasonably good. In summary, we conclude that aqueous glutaraldehyde consists primarily of the cyclic hemihydrate, equally distributed between its two stereoisomeric forms; that the proportion of the linear dihydrate is about 25% as high, and the hemihydrate about 50% as high as previously proposed;3 and that, a t the temperatures originally cited (-40°)), the contribution of free aldehyde to the equilibrium mixture has been substantially underestimated.

a,

2 Q) a

Registry No.-I, 110-30-8; 11, 51052-01-8; 111, 51052-02-9; IVa, 51052-03-0; IVb, 51052-04-1.

20 -

References and Notes /

111 I

I

30"

40"

50"

60'

70O

Temperature ("C) Figure 3. Composition ( s o l i d curves, I3C n m r ) and percentage of free aldehyde groups (broken curve, pmr4) in aqueous 25% glutaraldehyde us. t e m p e r a t u r e . Free aldehyde fractions c a l c u l a t e d from

the composition are i n d i c a t e d by points x . These spin coupling effects are not so clear in the proton spectra in lower magnetic fields, since they require both that the anomeric protons be clearly resolved and

(1) Union Carbide Corp., Chemicais and Plastics, New York, N. Y. (2) F. M . Richards and J. R . Knowies, J . M o l . Biol., 37,231 (1968) (3) P. M . Hardy, A . C. Nicholls, and H . N . Rydon, Chem. Cornmun.. 565 (1969). ( 4 ) P. D. Sherman and Q W. Decker. Technology Series Pamphlet. Union Carbide Corp., Research and Development Department, South Charleston, W. Va. 25303, April 10, 1970. (5) Spectra were recorded on a Bruker HX-90 multinuclear spectrometer locked on the 13F signal of an external wF6 coaxial capillary. with broad band decoupilng at 90 MHz, and a nominal observing frequency of 22.63 MHz, operating in the repetitively pulsed mode (pulse width = 5 psec, -30") with a acquisition rate of 100 p s e c i point 4K data points, a 1.0-sec repetition rate (0.6-sec delay), and a 100-psec RC filter. The filter arrangement causes a -60% attenuation of the highest frequencies, at the extreme right of the transformed spectrum in Figure 1 . intensity data were corrected for this effect using an empirically determined calibration curve. (6) J. B. Stothers, "Carbon-13 NMR Spectroscopy," Academic Press, New York, N . Y . , 1972, and original references cited therein. (7) J. D. Roberts, F J. Weigert, J. I . Kroschwitz, and H . J. Reich, J . Arner. Chern. SOC.,92, 1338 (1970),